Limits...
Transcription termination and chimeric RNA formation controlled by Arabidopsis thaliana FPA.

Duc C, Sherstnev A, Cole C, Barton GJ, Simpson GG - PLoS Genet. (2013)

Bottom Line: We define intergenic read-through transcripts resulting from defective RNA 3' end formation in fpa mutants and detail cryptic splicing and antisense transcription associated with these read-through RNAs.Finally, we show that defective termination at specific loci in fpa mutants is shared with dicer-like 1 (dcl1) or dcl4 mutants, leading us to develop alternative explanations for some silencing roles of these proteins.We relate our findings to the impact that altered patterns of 3' end formation can have on gene and genome organisation.

View Article: PubMed Central - PubMed

Affiliation: College of Life Sciences, University of Dundee, Dundee, Scotland, United Kingdom.

ABSTRACT
Alternative cleavage and polyadenylation influence the coding and regulatory potential of mRNAs and where transcription termination occurs. Although widespread, few regulators of this process are known. The Arabidopsis thaliana protein FPA is a rare example of a trans-acting regulator of poly(A) site choice. Analysing fpa mutants therefore provides an opportunity to reveal generic consequences of disrupting this process. We used direct RNA sequencing to quantify shifts in RNA 3' formation in fpa mutants. Here we show that specific chimeric RNAs formed between the exons of otherwise separate genes are a striking consequence of loss of FPA function. We define intergenic read-through transcripts resulting from defective RNA 3' end formation in fpa mutants and detail cryptic splicing and antisense transcription associated with these read-through RNAs. We identify alternative polyadenylation within introns that is sensitive to FPA and show FPA-dependent shifts in IBM1 poly(A) site selection that differ from those recently defined in mutants defective in intragenic heterochromatin and DNA methylation. Finally, we show that defective termination at specific loci in fpa mutants is shared with dicer-like 1 (dcl1) or dcl4 mutants, leading us to develop alternative explanations for some silencing roles of these proteins. We relate our findings to the impact that altered patterns of 3' end formation can have on gene and genome organisation.

Show MeSH

Related in: MedlinePlus

Differentially expressed transposons between wild-type and fpa-7.(A) Differential expression of the transposable element gene (At5g10670) in fpa-7. (B) Read-through contiguous RNAs were validated by RT-PCR (red dashed line). Three biological replicates (1, 2 and 3) were used for each genotype: wild-type (WT) and fpa-7. UBIQUITIN LIGASE 21 (UBC21) was used as a control. RT-PCR products were separated on agarose gels and stained with ethidium bromide. (C) Transcripts are either cleaved and polyadenylated in the annotated 3′UTR or at the intergenic sites, as determined by sequencing the cloned RT-PCR products. Red rectangles represent the 3′UTR specific to the read-through transcript and red lines represent 3′UTR introns. (D) Differential expression of the transposable element gene (At5g35935) in fpa-7. Recent re-annotation of At5g35935 [12], [13], [28], [29] defines two transcription units within it: the recently arisen pseudogene psORF and the transposon At5TE50260. DRS data reveal that silencing of psORF is lost in fpa-7. (E) RT-qPCR analysis of psORF in fpa-7 and fpa-8 mutant alleles. Silencing of psORF (p2 and p2b) is lost in fpa-7 but not in fpa-8. Data are the means ± SEM obtained for three independent PCR amplifications of three biological replicates. The y-axis shows the fold change relative to WT (WT set to 1) after normalisation to UBC21 gene expression. Location of the RT-qPCR amplicon is displayed on the left panel. *, P<0.05; Student's t-test. Normalised reads mapping to the different loci are presented for WT and fpa. Genes are orientated 5′–3′; exons are denoted by rectangles, UTRs by adjoining narrower rectangles and introns by lines. Images of normalised read alignments were made using the Integrated Genome Browser [55] and correspond to combined reads from the three sequenced biological replicates for each genotype.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3814327&req=5

pgen-1003867-g003: Differentially expressed transposons between wild-type and fpa-7.(A) Differential expression of the transposable element gene (At5g10670) in fpa-7. (B) Read-through contiguous RNAs were validated by RT-PCR (red dashed line). Three biological replicates (1, 2 and 3) were used for each genotype: wild-type (WT) and fpa-7. UBIQUITIN LIGASE 21 (UBC21) was used as a control. RT-PCR products were separated on agarose gels and stained with ethidium bromide. (C) Transcripts are either cleaved and polyadenylated in the annotated 3′UTR or at the intergenic sites, as determined by sequencing the cloned RT-PCR products. Red rectangles represent the 3′UTR specific to the read-through transcript and red lines represent 3′UTR introns. (D) Differential expression of the transposable element gene (At5g35935) in fpa-7. Recent re-annotation of At5g35935 [12], [13], [28], [29] defines two transcription units within it: the recently arisen pseudogene psORF and the transposon At5TE50260. DRS data reveal that silencing of psORF is lost in fpa-7. (E) RT-qPCR analysis of psORF in fpa-7 and fpa-8 mutant alleles. Silencing of psORF (p2 and p2b) is lost in fpa-7 but not in fpa-8. Data are the means ± SEM obtained for three independent PCR amplifications of three biological replicates. The y-axis shows the fold change relative to WT (WT set to 1) after normalisation to UBC21 gene expression. Location of the RT-qPCR amplicon is displayed on the left panel. *, P<0.05; Student's t-test. Normalised reads mapping to the different loci are presented for WT and fpa. Genes are orientated 5′–3′; exons are denoted by rectangles, UTRs by adjoining narrower rectangles and introns by lines. Images of normalised read alignments were made using the Integrated Genome Browser [55] and correspond to combined reads from the three sequenced biological replicates for each genotype.

Mentions: Although FPA has been reported to play a widespread role in RNA-mediated chromatin silencing [21], we found statistically significant increases in fpa-7 DRS counts at only 28 of the 31,189 transposons and 3,903 transposable element genes annotated in TAIR10 (Table S3). Since we had previously found that the apparent involvement of FPA in silencing the SINE retroelement AtSN1 could be explained by read-through resulting from defective termination at an upstream Pol II gene [14], [21], we asked whether reads mapping to annotated transposons reflect a genuine loss of silencing or whether they could also be explained by read-through. We found, for example, that DRS reads mapping to an apparently up-regulated transposable element gene (At5g10670) did indeed result from read-through from the upstream protein-coding gene At5g10690 (Figure 3A–C). Therefore, at least some of the relatively small number of reads mapping to transposons in this study may also be explained by read-through events. Clearly, not all misregulated transposons will be polyadenylated and so they will not be detected here, but the results of this genome-wide analysis are inconsistent with the suggestion that FPA plays a widespread role in RNA-mediated chromatin silencing. This conclusion is supported by a recent DNA methylation analysis of A. thaliana silencing mutants that included fpa-7 and found no evidence of FPA affecting RNA-dependent DNA methylation (RdDM) target sites [27].


Transcription termination and chimeric RNA formation controlled by Arabidopsis thaliana FPA.

Duc C, Sherstnev A, Cole C, Barton GJ, Simpson GG - PLoS Genet. (2013)

Differentially expressed transposons between wild-type and fpa-7.(A) Differential expression of the transposable element gene (At5g10670) in fpa-7. (B) Read-through contiguous RNAs were validated by RT-PCR (red dashed line). Three biological replicates (1, 2 and 3) were used for each genotype: wild-type (WT) and fpa-7. UBIQUITIN LIGASE 21 (UBC21) was used as a control. RT-PCR products were separated on agarose gels and stained with ethidium bromide. (C) Transcripts are either cleaved and polyadenylated in the annotated 3′UTR or at the intergenic sites, as determined by sequencing the cloned RT-PCR products. Red rectangles represent the 3′UTR specific to the read-through transcript and red lines represent 3′UTR introns. (D) Differential expression of the transposable element gene (At5g35935) in fpa-7. Recent re-annotation of At5g35935 [12], [13], [28], [29] defines two transcription units within it: the recently arisen pseudogene psORF and the transposon At5TE50260. DRS data reveal that silencing of psORF is lost in fpa-7. (E) RT-qPCR analysis of psORF in fpa-7 and fpa-8 mutant alleles. Silencing of psORF (p2 and p2b) is lost in fpa-7 but not in fpa-8. Data are the means ± SEM obtained for three independent PCR amplifications of three biological replicates. The y-axis shows the fold change relative to WT (WT set to 1) after normalisation to UBC21 gene expression. Location of the RT-qPCR amplicon is displayed on the left panel. *, P<0.05; Student's t-test. Normalised reads mapping to the different loci are presented for WT and fpa. Genes are orientated 5′–3′; exons are denoted by rectangles, UTRs by adjoining narrower rectangles and introns by lines. Images of normalised read alignments were made using the Integrated Genome Browser [55] and correspond to combined reads from the three sequenced biological replicates for each genotype.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3814327&req=5

pgen-1003867-g003: Differentially expressed transposons between wild-type and fpa-7.(A) Differential expression of the transposable element gene (At5g10670) in fpa-7. (B) Read-through contiguous RNAs were validated by RT-PCR (red dashed line). Three biological replicates (1, 2 and 3) were used for each genotype: wild-type (WT) and fpa-7. UBIQUITIN LIGASE 21 (UBC21) was used as a control. RT-PCR products were separated on agarose gels and stained with ethidium bromide. (C) Transcripts are either cleaved and polyadenylated in the annotated 3′UTR or at the intergenic sites, as determined by sequencing the cloned RT-PCR products. Red rectangles represent the 3′UTR specific to the read-through transcript and red lines represent 3′UTR introns. (D) Differential expression of the transposable element gene (At5g35935) in fpa-7. Recent re-annotation of At5g35935 [12], [13], [28], [29] defines two transcription units within it: the recently arisen pseudogene psORF and the transposon At5TE50260. DRS data reveal that silencing of psORF is lost in fpa-7. (E) RT-qPCR analysis of psORF in fpa-7 and fpa-8 mutant alleles. Silencing of psORF (p2 and p2b) is lost in fpa-7 but not in fpa-8. Data are the means ± SEM obtained for three independent PCR amplifications of three biological replicates. The y-axis shows the fold change relative to WT (WT set to 1) after normalisation to UBC21 gene expression. Location of the RT-qPCR amplicon is displayed on the left panel. *, P<0.05; Student's t-test. Normalised reads mapping to the different loci are presented for WT and fpa. Genes are orientated 5′–3′; exons are denoted by rectangles, UTRs by adjoining narrower rectangles and introns by lines. Images of normalised read alignments were made using the Integrated Genome Browser [55] and correspond to combined reads from the three sequenced biological replicates for each genotype.
Mentions: Although FPA has been reported to play a widespread role in RNA-mediated chromatin silencing [21], we found statistically significant increases in fpa-7 DRS counts at only 28 of the 31,189 transposons and 3,903 transposable element genes annotated in TAIR10 (Table S3). Since we had previously found that the apparent involvement of FPA in silencing the SINE retroelement AtSN1 could be explained by read-through resulting from defective termination at an upstream Pol II gene [14], [21], we asked whether reads mapping to annotated transposons reflect a genuine loss of silencing or whether they could also be explained by read-through. We found, for example, that DRS reads mapping to an apparently up-regulated transposable element gene (At5g10670) did indeed result from read-through from the upstream protein-coding gene At5g10690 (Figure 3A–C). Therefore, at least some of the relatively small number of reads mapping to transposons in this study may also be explained by read-through events. Clearly, not all misregulated transposons will be polyadenylated and so they will not be detected here, but the results of this genome-wide analysis are inconsistent with the suggestion that FPA plays a widespread role in RNA-mediated chromatin silencing. This conclusion is supported by a recent DNA methylation analysis of A. thaliana silencing mutants that included fpa-7 and found no evidence of FPA affecting RNA-dependent DNA methylation (RdDM) target sites [27].

Bottom Line: We define intergenic read-through transcripts resulting from defective RNA 3' end formation in fpa mutants and detail cryptic splicing and antisense transcription associated with these read-through RNAs.Finally, we show that defective termination at specific loci in fpa mutants is shared with dicer-like 1 (dcl1) or dcl4 mutants, leading us to develop alternative explanations for some silencing roles of these proteins.We relate our findings to the impact that altered patterns of 3' end formation can have on gene and genome organisation.

View Article: PubMed Central - PubMed

Affiliation: College of Life Sciences, University of Dundee, Dundee, Scotland, United Kingdom.

ABSTRACT
Alternative cleavage and polyadenylation influence the coding and regulatory potential of mRNAs and where transcription termination occurs. Although widespread, few regulators of this process are known. The Arabidopsis thaliana protein FPA is a rare example of a trans-acting regulator of poly(A) site choice. Analysing fpa mutants therefore provides an opportunity to reveal generic consequences of disrupting this process. We used direct RNA sequencing to quantify shifts in RNA 3' formation in fpa mutants. Here we show that specific chimeric RNAs formed between the exons of otherwise separate genes are a striking consequence of loss of FPA function. We define intergenic read-through transcripts resulting from defective RNA 3' end formation in fpa mutants and detail cryptic splicing and antisense transcription associated with these read-through RNAs. We identify alternative polyadenylation within introns that is sensitive to FPA and show FPA-dependent shifts in IBM1 poly(A) site selection that differ from those recently defined in mutants defective in intragenic heterochromatin and DNA methylation. Finally, we show that defective termination at specific loci in fpa mutants is shared with dicer-like 1 (dcl1) or dcl4 mutants, leading us to develop alternative explanations for some silencing roles of these proteins. We relate our findings to the impact that altered patterns of 3' end formation can have on gene and genome organisation.

Show MeSH
Related in: MedlinePlus